Novel high performance poly(p-phenylene benzobisimidazole) (PBDI) membranes fabricated by interfacial polymerization for H 2 separation

(1)

Novel high performance poly(p-phenylene benzobisimidazole) (PBDI) membranes

fabricated by interfacial polymerization for H 2 separation

Shan, Meixia; Liu, Xinlei; Wang, Xuerui; Liu, Zilong; Iziyi, Hodayfa; Ganapathy, Swapna; Gascon, Jorge;

Kapteijn, Freek

DOI

10.1039/c9ta01524h

Publication date

2019

Document Version

Accepted author manuscript

Published in

Journal of Materials Chemistry A

Citation (APA)

Shan, M., Liu, X., Wang, X., Liu, Z., Iziyi, H., Ganapathy, S., Gascon, J., & Kapteijn, F. (2019). Novel high

performance poly(p-phenylene benzobisimidazole) (PBDI) membranes fabricated by interfacial

polymerization for H 2 separation. Journal of Materials Chemistry A, 7(15), 8929-8937.

https://doi.org/10.1039/c9ta01524h

Important note

To cite this publication, please use the final published version (if applicable).

Please check the document version above.

Copyright

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons. Takedown policy

Please contact us and provide details if you believe this document breaches copyrights. We will remove access to the work immediately and investigate your claim.

(2)

This is an Accepted Manuscript, which has been through the

Royal Society of Chemistry peer review process and has been

accepted for publication.

Accepted Manuscripts are published online shortly after

acceptance, before technical editing, formatting and proof reading.

Using this free service, authors can make their results available

to the community, in citable form, before we publish the edited

article. We will replace this Accepted Manuscript with the edited

and formatted Advance Article as soon as it is available.

You can find more information about Accepted Manuscripts in the

author guidelines

.

Please note that technical editing may introduce minor changes

to the text and/or graphics, which may alter content. The journal’s

standard

Terms & Conditions

and the ethical guidelines, outlined

in our

author and reviewer resource centre

, still apply. In no

event shall the Royal Society of Chemistry be held responsible

for any errors or omissions in this Accepted Manuscript or any

consequences arising from the use of any information it contains.

Accepted Manuscript

Journal of

Materials Chemistry A

Materials for energy and sustainability

www.rsc.org/MaterialsA

ISSN 2050-7488

Volume 4 Number 1 7 January 2016 Pages 1–330

PAPER Kun Chang, Zhaorong Chang et al. Bubble-template-assisted synthesis of hollow fullerene-like MoS2 nanocages as a lithium ion battery anode material

Journal of

Materials Chemistry A

Materials for energy and sustainability

This article can be cited before page numbers have been issued, to do this please use: M. Shan, X. Liu, X. Wang, Z. Liu, H. Iziyi, S. Ganapathy, J. Gascon and F. Kapteijn, J. Mater. Chem. A, 2019, DOI:

(3)

Journal Name

ARTICLE

a._{Catalysis Engineering, Chemical Engineering Department, Delft University of}

Technology, Van der Maasweg 9, 2629 HZ Delft, The Netherlands *Emails: x.liu-8@tudelft.nl; f.kapteijn@tudelft.nl

b_{Organic Materials & Interface, Chemical Engineering, Delft University of}

Technology, Van der Maasweg 9, 2629 HZ Delft, The Netherlands

c_{Department of Radiation Science and Technology, Delft University of Technology,}

Mekelweg 15, 2629 JB Delft, The Netherlands

d_{King Abdullah University of Science and Technology, KAUST Catalysis Center,}

Advanced Catalytic Materials. Thuwal 23955, Saudi Arabia Electronic Supplementary Information (ESI) available.

Received 00th January 20xx, Accepted 00th January 20xx

DOI: 10.1039/x0xx00000x

www.rsc.org/

Novel high performance poly(p-phenylene benzobisimidazole) (PBDI)

membranes fabricated by interfacial polymerization for H

2

separation

Meixia Shana_{, Xinlei Liu}a*_{, Xuerui Wang}a_{, Zilong Liu}b_{, Hodayfa Iziyi}a_{, Swapna Ganapathy}c_{, Jorge}

Gascona,d_{and Freek Kapteijn}a*

Membranes with high selectivity and permeance are needed to reduce the energy consumption in hydrogen purification and pre-combustion CO2 capture. Polybenzimidazole (PBI) is one of the leading membrane materials for this separation. In

this study, we present superior novel supported PBI (poly(p-phenylene benzobisimidazole, PBDI) membranes prepared by a facile interfacial polymerization (IP) method. The effect of IP reaction duration, operating temperature and pressure on membrane separation performance was systematically investigated. The best performance was achieved for membranes prepared in 2 h reaction time. The resulting membranes display an ultrahigh mixed-gas H2/CO2 selectivity of 23 at 423 K

together with an excellent H2 permeance of 241 GPU, surpassing the membrane performance of conventional polymers

(the 2008 Robeson upper bound). These separation results, together with the facile manufacture, pressure resistance, long-term thermostability ( > 200 h) and economic analysis, recommend the PBDI membranes for industrial use in H2

purification and pre-combustion CO2 capture. Besides, PBDI membranes possesss high selectivities towards H2/N2 (up to

60) and H2/CH4 (up to 48) mixtures, indicating their potential applications in ammonia synthesis and syngas production.

1. Introduction

The consumption of fossil fuels has brought a variety of environmental issues. Hydrogen holds great potential as an alternative fuel because of its high combustion heat and zero-carbon emission (only water as product). 1, 2_{Besides, H}

2 is also

widely used in the chemical industry for hydrogenation reactions,3

methanol and ammonia production,4_{which makes it an important}

commodity. Currently, approximately 80 % of the hydrogen is produced by natural gas and petroleum reforming.5_{Production of}

H2 through abundant coal or biomass via gasification is also a viable

alternative.6_{However, the main by-product of these processes is}

CO2, which is a major environmental concern due to its

accompanying greenhouse effect. Thus, H2 purification by capturing

CO2 is drawing a lot of attention. Further, recovery of valuable H2

from industrial vent gases and adjusting H2/CO-CO2 ratio in syngas

are important industrial activities. The current technologies for hydrogen purification, such as scrubbing technology, pressure swing

adsorption and cryogenic distillation are highly energy and capital intensive.7_{Membranes that are H}₂_{-permeating and CO}₂_{-rejecting at}

syngas operating conditions (423 K and above) would provide a low-cost and energy-efficient routes for H2 purification and CO2

capture.8-12

Inorganic membranes, such as palladium,13_zeolites,14_graphene

oxide (GO),15_{and metal organic frameworks (MOFs)}16, 17_have

exhibited outstanding selectivity or permeance for H2/CO2

separation. However, high production costs, mechanical brittleness and the fabrication complexities make these inorganic membranes less commercially attractive. In addition, engineering these materials into thin membranes and on a large-scale is extremely challenging. Polymer membranes, on the other hand, are economically acceptable, easy to process and able to produce on a large scale.18

Gas permeation through a dense polymer membrane generally follows the solution-diffusion mechanism,19_{where gas permeability}

(P) is determined by both gas solubility (S) and diffusivity (D) (P = S ×

D). The smaller H2 molecule and the higher solubility of CO2 result in

an opposite trend in diffusivity and solubility selectivities, thus conventional polymer membranes usually have moderate H2/CO2

selectivity. For example, the commercial cellulose, polysulfone and Matrimid®_{polymers only show an H}₂_/CO₂_{selectivity around 3 at}

308 K.20,21_{In addition, the main drawback of polymeric membranes}

for H2/CO2 separation is the lack of thermochemical stability at

syngas operating conditions.

Journal

of

Materials

Chemistry

A

Accepted

Manuscript

Published on 14 March 2019. Downloaded on 3/15/2019 6:55:42 AM.

View Article Online

(4)

Polybenzimidazoles (PBI),22_{are a class of heterocyclic polymers that}

can meet the syngas separation requirements in theory because of their exceptional thermal and chemical stability and high intrinsic H2/CO2 selectivity.23 Unfortunately, PBI membrane exhibits very low

H2 permeability owing to the microstructural rigidity and tightly

chain packing. Several approaches have been taken to enhance PBI membrane permeability including polymer blending,24_structural

modification,25, 26 _chemical _{cross-linking,}27, 28 _thermal

rearrangement29_{and incorporation of fillers.}2, 30, 31_Besides

free-standing PBI membranes, asymmetric PBI membranes in the configuration of flat sheets and hollow fibers8, 25, 30, 32_{have been}

unveiled, prepared by phase inversion. Despite great achievements made in the development of thin PBI membranes, coating of an additional layer of silicone rubber was always applied to heal the defects and compensate the low processability of PBI.33_{In this}

sense, the development of novel PBI membrane materials and approaches for membrane fabrication are still highly on demand. Interfacial polymerization (IP) is a commercially common, well adapted technique for the synthesis of large-scale and ultrathin polyamide membranes for reverse osmosis and nanofilitration.34-37

Despite IP approach has been used in industry, it remains largely unexplored for preparing other polymeric membranes, such as PBI membranes. PBI was generally synthesized either by the coupling of phenylenediamines and carboxylic acids under strong acidic conditions and high temperatures or through the condensation reaction of phenylenediamines and aldehydes using some oxidative reagents,38_{resulting in intractable powders, which are nearly}

insoluble in common organic solvents, restricting their further processing into membranes.

In this work, we demonstrate the feasibility for fabricating thin novel PBI films and membranes on the basis of poly(p-phenylene benzobisimidazole (PBDI) materials,39, 40_{by using a room}

temperature IP protocol. The resulting membranes have been tested for H2/CO2 separation over a wide range of temperatures. In

addition, the influence of reacting duration, feed pressures, operating temperature and other gases (N2 and CH4) on the

membrane performance have also been investigated. As far as we know, this is the first report of using IP method for creating thin PBDI membranes with ultrahigh H2 separation performance (H2

permeance up to 241 GPU together with high H2/CO2 selectivity of

23 at 423 K).

2. Experimental

2.1 Materials

Terephthalaldehyde (TPA) (99%), 1,2,4,5-Benzenetetramine tetrahydrochloride (BTA), Toluene (anhydrous, 99.8%), Ethanol (95%), and N,N-Dimethylformamide (DMF, anhydrous, 99.8%) were purchased from Sigma Aldrich. Deionized water was supplied by Applied Sciences of Delft University of Technology. Asymmetric α-Al2O3 substrates (α-Al2O3, average pore size ~ 2.5 μm; γ-Al2O3 layer

on top, average pore size ~ 5.0 nm) with a diameter of 18 mm and a thickness of 1 mm were purchased from Fraunhofer-Institut für

Keramische Technologien und Systeme IKTS. All materials and solvents were used without further purification.

2.2. Synthesis of PBDI films

A TPA solution in toluene (1.0 wt.%) was gently spread on top of a 1.5 wt.% aqueous solution containing BTA in a petri dish (or bottle). After a few seconds, a brown layer was formed at the water-toluene interface. The film was left for 3 h and then washed. The resulting films were left to dry in the fume hood and finally dried under vacuum at 373 K overnight. Plenty of films were prepared in this way for later characterization.

2.3 Synthesis of supported PBDI membranes

An α-Al2O3 substrate was first immersed in a small clean petri dish

(diameter ~ 5 cm) containing 1.5 wt.% BTA aqueous solution under reduced pressure of 0.2 bara (absolute pressure). After 20 min, the solution was taken out from the petri dish and the substrate was dried with compressed air until no visible droplets were left on the surface. The 1.0 wt.% TPA solution was then poured onto the BTA saturated α-Al2O3 substrate and the reaction was let to continue for

1, 2 or 3 h. The membrane was left in the fume hood overnight, washed, dried at RT and finally put into a homemade permeation setup for performance testing. The specific membrane preparation conditions are provided in Table 1. All membranes were prepared in duplicate to confirm the reproducibility (Table S1).

Table 1. Summary of membrane preparation conditions.

Membranes

Interfacial polymerization conditions Aqueous amine phase (wt.%) Aldehyde in toluene phase (wt.%) IP time (h) M1 1.5 1.0 1 M2 1.5 1.0 2 M3 1.5 1.0 3 2.4. Characterization

The thermostability was determined by thermogravimetric analysis (TGA) performed in a Mettler Toledo TGA/SDTA851e apparatus by measuring the mass loss of the sample while heating the sample under N2 (100 mL∙min-1) from 303 to

1073 K at a heating rate of 5 K·min-1_.

Diffuse reflectance infrared Fourier transform spectra (DRIFTs) of the PBDI films were acquired using a Nicolet 8700 FT-IR (Thermo Scientific) equipped with a high-temperature cell with CaF2 windows. The samples were preheated in He at 423 K for

30 min to eliminate the influence from the trapped solvents. A Bruker-D8 advanced diffractometer with Co-Kα radiation (λ =

1.78897 Å) was used to record the powder X-ray diffraction (PXRD) patterns of PBDI films. The samples were scanned in the 2θ range of 5o_{- 80}o_{using a scan speed of 0.2 seconds per step and a step size of}

0.02o_{in a continuous scanning mode.}

Scanning electron microscope (SEM) images of PBDI films were acquired with a JEOL JSM-6010LA InTouchScope microscope. The

Journal

of

Materials

Chemistry

A

Accepted

Manuscript

View Article Online

(5)

Journal Name ARTICLE

surface and cross-section images of the supported membranes were obtained using a DualBeam Strata 235 microscope (FEI) and an AURIGA Compact (Zeiss) microscope. Prior to the analyses, the specimens were sputter-coated with gold.

CO2 (273, 298 and 323 K) and N2 (77 K) adsorption/desorption

isotherms were recorded in a TriStar II 3020 (Micromeritics) instrument. Prior to the gas adsorption measurements, the samples were degassed at 423 K under N2 flow overnight.

Atomic Force Microscopy (AFM) tomography images were acquired in tapping mode using a silicon tip (NSG03, NT-MDT), with a nominal value of the tip radius of 7 nm and a nominal spring constant of 0.4–2.7 N/m. Images were recorded over a 1 μm × 1 μm area with 512 × 512 data points, at room temperature, and at a scan rate of 1 Hz.

Solid state Nuclear Magnetic Resonance (NMR) measurements were performed using a Bruker Ascend 500 spectrometer operating at a 13_{C frequency of 125.76 MHz and equipped with a two channel}

4 mm MAS probe head (Bruker) at a 10 kHz spinning speed. One dimensional 13_{C CP/MAS spectra were recorded, where a proton}

π/2 pulse length of 3 s and a cross polarization period of 2 ms was used. 140000 scans were accumulated with a recycle delay of 5 s between scans. An exponential apodization function corresponding to a line broadening of 80 Hz was applied prior to Fourier transformation.

2.5. Gas permeation experiments

The PBDI membranes (effective area of 1.33 cm2_{) were mounted in}

a flange between Viton® O-rings. This flange fits in a permeation module that was placed inside an oven in a home-made permeation setup.41_{The mixed gas permeation measurements were carried out}

using an equimolar mixture of H2 and CO2, N2 or CH4 (50 mL·min-1

for each gas) as feed. Helium (4.6 mL·min-1_{) was used as sweep gas}

at the permeate side. The absolute pressure of the feed stream was adjusted in a range of 1 - 5 bara through a back-pressure controller at the retentate side, keeping the permeate side at atmospheric pressure. The permeation module was placed in a convection oven to control the temperature from 298 to 423 K. An on-line gas chromatograph (Interscience Compact GC) equipped with a packed Carboxen 1010 PLOT (30 m x 0.32 mm) column and TCD and FID detectors was used to analyze the gas compositions in the permeate.

Gas separation performance is defined by the selectivity (α) and the gas permeance (P) of the individual components. The permeance of component i (Pi) was calculated as follows (Equation 1):

i i i i i N F P p A p = =    (1)

where Ni denotes the permeation rate of component i (mol s-1), Fi is

the molar flux of compound i (mol m-2_s-1_{), A is the membrane area}

and Δpi represents the partial pressure difference of component i

across the membrane and calculated according to Equation 2.

, ,

i feed i feed perm i perm

p p Y p X

 =  −  (2)

where pfeed and pperm represent the pressures at the feed and

permeate sides and Yi, feed and Xi, perm are the molar fractions of

component i in the feed and permeate gas streams, respectively.

The SI unit for the permeance is mol·s-1_·m-2_·Pa-1_{. However, gas}

permeances are widely reported in units GPU (Gas Permeation Unit), where 1 GPU = 3.35 x 10-10_mol·s-1_·m-2_·Pa-1_.

Gas permeability is an intrinsic property of the membrane material and typically used for the evaluation of the separation performance of dense polymeric membranes. The unit of permeability is Barrer, where 1 Barrer = 3.35 x 10-16_mol·s-1_·m-1_·Pa-1_{. Gas selectivity (α) was}

calculated as the ratio of the permeance of the faster permeating component (H2 in this study) to the permeance of the less

permeating component (CO2, N2 and CH4 in this study). For H2 and

CO2 as an example this reads (equation 3):

a_H 2/CO2 = PH2 P_CO 2 (3)

Ideal selectivity was applied in the case of single component data.

3. Results and discussion

3.1 Characterization

In order to study the feasibility of the IP protocol to fabricate PBDI membranes, as preliminary experiments the preparation of free-standing PBDI films was conducted at the interface of two immiscible liquid phases (Fig. 1). As depicted in Fig. 1a, a brown layer formed at the interface. The amine groups in aqueous BTA solution reacted quickly with the aldehyde moieties in toluene upon contacting. The growing film itself behaves as a barrier and confines the reaction to the remaining defects, yielding a continuous film (see Movie S1). The film surface was smooth with some spherical particles on top (Fig. 2c), attributed to emulsion droplets near the organic interface. A large number of spherical particles were observed beneath the film (Fig. 2d) because of the hydrophilicity of the PBDI.34_{The dense film layer has a thickness of around 300 nm}

(Fig. 2a and b).

Fig. 1 (a) Photograph of a free-standing PBDI film formed at the toluene-water interface. (b) Description of PBDI film preparation procedure. The aqueous BTA solution (bottom purple area) was contacted with TFB toluene solution (upper grey part), enabling the formation of PBDI film (orange part) at the interface. (c) Reaction scheme for PBDI.

The formation of benzimidazoles was confirmed by DRIFT spectroscopy and 13_{CP/MAS NMR. (Fig. 3a and b). Specifically, the}

C=N stretching band of the benzimidazole system in the vicinity of

Journal

of

Materials

Chemistry

A

Accepted

Manuscript

View Article Online

(6)

Fig. 2 Cross-sectional (a and b) and surface SEM images of a free-standing PBDI film (c, surface faced toluene phase, and d, surface faced water phase).

1616 cm-1_{together with the skeleton vibration of benzimidazole}

ring at 1490, 1375, and 1240 cm-1_{were observed. In line with the}

DRIFTs observations, the successful condensation between BTA and TPA units were further corroborated by 13_{CP/MAS NMR. The full}

assignments of the peaks are shown in Fig. 3b. Specifically, a signal at 151 ppm corresponds to the characteristic stretching of N=C-N band in the benzimidazole ring. Additional signals at 100 and 130 ppm are characteristic of aromatic carbons, indicating the incorporation of BTA and TPA units in the network. The thermal stability of the film was evaluated by TGA. PBDI is stable up to 800 K (Fig. 3c). The slight weight loss below 390 K can be attributed to the desorption of moisture or solvent trapped in the network. The weight loss between 390 K and 640 K was caused by the decomposition of oligomer and residual monomers trapped in the network.42

Furthermore, the X-ray diffraction pattern of PBDI films presented in Fig. 3d reflects a diffraction pattern different from normal polybenzimidazoles. One diffraction assigned to the d-spacing of 0.4 nm was reported for PBI membranes28_{whereas two diffractions at}

13° and 30°, corresponding to a d-spacing of 0.79 nm and 0.35 nm, respectively, were observed for PBDI, demonstrating a different packing of polymer chains. The two reflections of PBDI can be probably assigned to the distance of polybenzimidazole chains and the parallel orientation of benzimidazole rings with respect to the film surface.43, 44

In order to study the texture properties of the prepared film, low-pressure N2 sorption measurement was performed at 77 K (Fig. 3e),

The film shows a very low N2 uptake and the isotherms are fully

reversible, indicating the low accessibility of the network towards N2, pointing at a submicroporous material. The

Brunauer-Emmett-Teller area was too low to be calculated, reached the detection limit of the adsorption instrument. Furthermore, CO2 adsorption

isotherms at different temperatures (Fig. 3f) were determined to check whether diffusion limitations at 77 K could be the reason for the small N2 uptake. The low CO2 adsorption further confirms the

dense packing structure of the PBDI film. Moreover, the Virial equation (Eq. (S1), (S2), see details in SI) was used to calculate the steric heat of CO2 adsorption 𝑸𝒔𝒕 on PBDI, ranging from ~34 kJ mol-1

(Fig. S1) at low loading, which is even slightly higher than for other benzimidazole polymers,45, 46_{indicating its high binding ability}

towards CO2, decreasing to ~30 kJ mol-1 at higher loading.

Compared with the free-standing film, the membrane formed on top of the alumina support is relatively thicker, around 800 nm (Fig. 4d). It seems that the solid substrate effectively anchored the membrane, preventing the migration of PBDI oligomers to the aqueous phase and stabilizing the interface for continuous reaction. Particles with size down to 20 nm were possiblyrecognized on the

Journal

of

Materials

Chemistry

A

Accepted

Manuscript

View Article Online

(7)

Fig. 3 Characterization of PBDI films: (a) DRIFT spectra, (b) 13_{C CP/MAS spectra, (c) TGA profile, (d) PXRD pattern of PBDI film, (e) N}

2 adsorption (closed

symbols)/ desorption (open symbols) isotherm at 77 K and (f) CO2 adsorption (closed symbols) / desorption (open symbols) isotherms at 273, 298 and 323 K.

Peaks labelled with * in b were assigned to the spinning sidebands. The inset in e is the photo of PBDI film pieces in a glass sample bottle.

Fig. 4. SEM (a) and AFM (b) images of the membrane surface. (c) and (d) cross-section SEM images of the composite membrane. Results are obtained from M2-1.

membrane surface from the SEM and AFM images (Fig. 4a, b and Fig. S2).

3.2. Gas permeation

The reproducibility of the preparation procedure of all the three membranes (with reaction duration of 1, 2 and 3 h) was confirmed by the H2/CO2 separation performance at 1 bara, 373 K (Table S1).

Permeation of single gas

Before gas permeation, the as-synthesized PBDI membranes (M3-1) were on-stream activated at 373 K until a steady state was reached. The membrane was tested for permeation using H2, CO2, N2, and

CH4 in single gas mode. The permeation results are provided in

Table 2 and Fig. 5. The order of gas permeance roughly depends on their kinetic diameters, and there is a clearly cut-off between H2

and the other larger gases (CO2, N2 and CH4). The ideal selectivity of

H2 over CO2, N2 and CH4 is 18.9, 59.7 and 47.5, respectively and

Journal

of

Materials

Chemistry

A

Accepted

Manuscript

View Article Online

(8)

exceeds by far the corresponding Knudsen selectivity (Table 2), reflecting the formation of a dense membrane. However, the permeance does not exactly follow the order of the kinetic diameters of the measured gas molecules as exemplified by the case of CH4. This is probably due to some preferable adsorption of

CH4 over N2 on PBDI. Similar adsorption results had been reported

for other imidazole-linked polymers.47, 48

Table 2. Permeances, ideal selectivities for single gas at 373 K and 1 bara feed pressure (M3-1 membrane).

Gas Kinetic diameter (Å) Permeance (GPU) Ideal selectivity (H2/-) Knudsen selectivity H2 2.89 71.7 - CO2 3.3 3.8 18.9 4.7 N2 3.64 1.2 59.7 3.7 CH4 3.76 1.5 47.5 2.8

Effect of reaction duration

Interfacial polymerization was performed at the monomer concentration of 1.5 and 1.0 wt.% for aqueous BTA and organic TPA phase, respectively. Fig. 6a presents the gas separation performance at 423 K of three membranes synthesized for 1, 2 and 3 h, respectively. Both gas permeance (H2 and CO2) and H2/CO2

selectivity increase with increasing the reaction time from 1 h to 2 h. We speculate that this is due to the fact that the unreacted monomers may block gas permeation after only 1 h reaction. Longer synthesis times would allow for more monomers to encounter the reactive functional groups, leaving more free volume for gas permeation. The gas permeance started to decrease when further increase reaction time to 3 h owing to the formation of

Fig. 5 Pure gas permeances of PBDI membrane at 373 K and ideal selectivity (M3-1 membrane).

additional layers thus increasing the thickness of the selective layer. As depicted in Fig. 4d and Fig. S4, the membrane thickness is ca. 0.60, 0.80 and 1.3 µm, corresponding to 1, 2 and 3 h reaction duration, respectively. The H2/CO2 selectivity seems to level off at

23 after 2 h reaction and the 2 h membrane could reach a permeance of 241 GPU.

Effect of feed temperature

Generally, gas transport in dense polymer membranes results from a combination of sorption and diffusion. H2 shows high diffusion in

polymer membranes owing to the smaller molecular size and CO2

exhibits higher sorption ability because of its high condensability

and polarity. The interplay between diffusion and sorption as a function of temperature can be optimized to obtain enhanced separation performance. Thus, single and mixed-gas permeation tests were performed at different temperatures to well understand the gas transport properties of PBDI membrane. Both the H2 and

CO2 permeance of the membrane increased with increasing

temperature (Fig. 6b), demonstrating an activated diffusion dominated gas transport. The effect of temperature on gas permeance (see details in SI) was captured by an Arrhenius relation. The apparent activation energy for H2 permeation is higher than for

CO2 in both mixed and single gas permeation due to the moderating

effect of the heat of adsorption for CO2 (Fig. S6 and Table S3). The

stronger influence of temperature on H2 permeation, results in an

increase in H2/CO2 selectivity with temperature (up to 373 K).

Effect of feed pressure

In addition to operating temperature, the effect of feed pressure (ranging from 1 to 5 bara) on both single and mixed-gas permeation were conducted at 373 K to further understand the gas transport behavior and pressure resistance of PBDI membranes. As depicted in Fig. 6c, both H2 and CO2 permeance decreased with feed pressure.

For CO2 this is usually attributed to an effect of non-linear

adsorption, but adsorption of H2 will be linear if any. It is

Journal

of

Materials

Chemistry

A

Accepted

Manuscript

View Article Online

(9)

Fig. 6 Effect of (a) reaction duration (permeation at 423 K, 1 bara), (b) temperature (at 1 bara) and (c) pressure (at 373 K) on H2/CO2 mixture separation

performance of PBDI membrane. Dashed and solid lines in (c) correspond to single component and equimolar mixed gas test, respectively. (d) Long-time stability for H2/CO2 separation under 373 and 423 K (blue shaded area) at 1 bara. Samples used M1-1, M2-1 and M3-1 (a), M2-1 (b), and M3-2 (c) and M3-1

(d).

interpreted that the higher feed pressure leads to a more compact packing of polymer chains.49_{Plasticization by CO}₂_{is therefore not}

observed in the studied pressure range. Compared with the single component test, H2 permeance and selectivity slightly dropped in

the mixed gas operation due to the competitive permeation of CO2.

The supported membrane was robust at 5 bara confirming a good mechanical strength.

Stability

High-performance polymeric membranes with good stability at elevated temperatures are required to realize a robust membrane system. From this point of view, the durability of PBDI membrane was examined at 373 and 423 K. The membrane can maintain the high permeance and selectivity with small fluctuation over 200 h (Fig. 6d), verifying its immutable structure and applicability for long-term use.

Comparison with literature

For comparison, the unveiled performance of membranes based on the PBI family in the configuration of flat sheet and hollow fiber are summarized in Table S4 and presented in Fig. 7a. The PBDI membranes are clearly superior to most of the state-of-the-art PBI membranes, highlighting the significance of IP protocol and the novel material. Furthermore, the potential of the PBDI membrane for real industrial application was assessed referring to the target performance region (Fig. 7a, green area). As reported,50, 51_{with a H}₂

permeance >100 GPU and H2/CO2 selectivity >15, the

environmental-friendly membrane technology can significantly reduce hydrogen production cost against the conventional cold absorption-based scrubbing processes. The performance of PBDI membrane (M2-1) with a H2 permeance of 241 GPU and selectivity

of 23, clearly located in the commercial target region, pointing out its value for real application.

Journal

of

Materials

Chemistry

A

Accepted

Manuscript

View Article Online

(10)

Fig. 7 (a) Comparison of H2 permeance and H2/CO2 selectivity of our PBDI membranes (M2-1 and M3-1) with previously reported PBI based membranes, as

well as the target performance region for pre-combustion CO2 capture (inset green area).50, 51 (b) Comparison of H2 permeability and H2/CO2 selectivity of our

membrane with previously reported polymer membranes. The 2008 Robeson bound line52_{was shown for reference. The data points were plotted referring}

to Table S4 and S5.

Additionally, in order to benchmark and give a general overview of the membrane material performance, the PBDI membrane thickness was used for normalization. The reported permeability of PBI and modified PBI membranes (instead of permeance) are listed (Table S5) and plotted together with the Robeson upper bound52

(Fig. 7b). The PBDI membrane performance transcended the Robeson upper bound by a wide margin and outperformed both the PBI and modified PBI membranes, demonstrating the developed PBDI membranes possess excellent H2 permeability and H2/CO2

selectivity.

4. Conclusions

In summary, free-standing PBDI films and supported PBDI membranes were successfully prepared for the first time through a facile, easy to scale up interfacial polymerization approach. Single component and mixture permeation data demonstrates a high H2

selectivity over CO2, N2 and CH4 of the membrane, reflecting the

formation of a dense layer. An optimal membrane was obtained for an interfacial polymerization reaction time of 2 h, possessing an ultrahigh hydrogen permeance of 241 GPU and a H2/CO2 selectivity

of 23 at 423 K, which overcomes the 2008 Robeson upper bound limit by a wide margin. In addition, the membrane performed stable during ~200 h at elevated temperatures and pressure. Results of an economic evaluation indicate that our membrane is located well in the target performance region and is more energy efficient than previously reported PBI membranes, demonstrating the great advantages of IP method for the preparation of defect-free PBI membranes. The appealing H2/CO2 separation performance, also

for H2 mixtures with N2 and CH4, together with such a simple

fabrication method paves a way for the industrial use of membranes for H2 purification and pre-combustion CO2 capture.

Conflicts of interest

The authors declare that they have no competing interests except one patent application (U.S. Provisional Patent NO. 62/659, 271)

Acknowledgements

Meixia Shan gratefully acknowledge the supporting from the China Scholarship Council.

Notes and References

1. J. A. Turner, Science, 2004, 305, 972-974.

2. T. Yang, Y. Xiao and T.-S. Chung, Energy & Environmental

Science, 2011, 4, 4171-4180.

3. R. Mariscal, P. Maireles-Torres, M. Ojeda, I. Sádaba and M. López Granados, Energy & Environmental Science, 2016, 9, 1144-1189.

4. A. Klerke, C. H. Christensen, J. K. Nørskov and T. Vegge,

Journal of Materials Chemistry, 2008, 18, 2304-2310.

5. G. J. Stiegel and M. Ramezan, International Journal of Coal

Geology, 2006, 65, 173-190.

6. H. Lin, E. Van Wagner, B. D. Freeman, L. G. Toy and R. P. Gupta, Science, 2006, 311, 639-642.

7. N. W. Ockwig and T. M. Nenoff, Chemical Reviews, 2007, 107, 4078-4110.

8. K. A. Berchtold, R. P. Singh, J. S. Young and K. W. Dudeck,

Journal of Membrane Science, 2012, 415-416, 265-270.

9. B. Seoane, J. Coronas, I. Gascon, M. E. Benavides, O. Karvan, J. Caro, F. Kapteijn and J. Gascon, Chemical

Society Reviews, 2015, 44, 2421-2454.

10. S. Wang, X. Li, H. Wu, Z. Tian, Q. Xin, G. He, D. Peng, S. Chen, Y. Yin, Z. Jiang and M. D. Guiver, Energy &

Environmental Science, 2016, 9, 1863-1890.

11. T.-H. Bae and J. R. Long, Energy & Environmental Science, 2013, 6, 3565-3569.

12. S. Japip, K.-S. Liao and T.-S. Chung, Advanced Materials, 2017, 29, 1603833.

Journal

of

Materials

Chemistry

A

Accepted

Manuscript

View Article Online

(11)

13. B. Zhu, C. H. Tang, H. Y. Xu, D. S. Su, J. Zhang and H. Li,

Journal of Membrane Science, 2017, 526, 138-146.

14. J. K. Das, N. Das and S. Bandyopadhyay, Journal of

Materials Chemistry A, 2013, 1, 4966-4973.

15. H. Li, Z. Song, X. Zhang, Y. Huang, S. Li, Y. Mao, H. J. Ploehn, Y. Bao and M. Yu, Science, 2013, 342, 95-98. 16. Y. Peng, Y. Li, Y. Ban, H. Jin, W. Jiao, X. Liu and W. Yang,

Science, 2014, 346, 1356-1359.

17. X. Wang, C. Chi, K. Zhang, Y. Qian, K. M. Gupta, Z. Kang, J. Jiang and D. Zhao, Nature Communications, 2017, 8, 14460.

18. Z. Qiao, S. Zhao, M. Sheng, J. Wang, S. Wang, Z. Wang, C. Zhong and M. D. Guiver, Nature Materials, 2018, DOI: 10.1038/s41563-018-0221-3.

19. J. G. Wijmans and R. W. Baker, Journal of Membrane

Science, 1995, 107, 1-21.

20. M. J. C. Ordoñez, K. J. Balkus, J. P. Ferraris and I. H. Musselman, Journal of Membrane Science, 2010, 361, 28-37.

21. D. F. Sanders, Z. P. Smith, R. Guo, L. M. Robeson, J. E. McGrath, D. R. Paul and B. D. Freeman, Polymer, 2013, 54, 4729-4761.

22. H. Vogel and C. S. Marvel, Journal of Polymer Science, 1961, 50, 511-539.

23. T. Yang and T.-S. Chung, International Journal of Hydrogen

Energy, 2013, 38, 229-239.

24. S. S. Hosseini, N. Peng and T. S. Chung, Journal of

Membrane Science, 2010, 349, 156-166.

25. S. C. Kumbharkar and K. Li, Journal of Membrane Science, 2012, 415-416, 793-800.

26. S. C. Kumbharkar and U. K. Kharul, Journal of Membrane

Science, 2010, 357, 134-142.

27. L. Zhu, M. T. Swihart and H. Lin, Journal of Materials

Chemistry A, 2017, 5, 19914-19923.

28. L. Zhu, M. T. Swihart and H. Lin, Energy & Environmental

Science, 2018, 11, 94-100.

29. S. H. Han, J. E. Lee, K.-J. Lee, H. B. Park and Y. M. Lee,

30. T. Yang, G. M. Shi and T.-S. Chung, Advanced Energy

Materials, 2012, 2, 1358-1367.

31. J. Sánchez-Laínez, B. Zornoza, S. Friebe, J. Caro, S. Cao, A. Sabetghadam, B. Seoane, J. Gascon, F. Kapteijn, C. Le Guillouzer, G. Clet, M. Daturi, C. Téllez and J. Coronas,

32. S. C. Kumbharkar, Y. Liu and K. Li, Journal of Membrane

Science, 2011, 375, 231-240.

33. J. Sánchez-Laínez, B. Zornoza, C. Téllez and J. Coronas,

34. S. Karan, Z. Jiang and A. G. Livingston, Science, 2015, 348, 1347-1351.

35. B.-H. Jeong, E. M. V. Hoek, Y. Yan, A. Subramani, X. Huang, G. Hurwitz, A. K. Ghosh and A. Jawor, Journal of

36. M. J. T. Raaijmakers and N. E. Benes, Progress in Polymer

Science, 2016, 63, 86-142.

37. M. F. Jimenez-Solomon, Q. Song, K. E. Jelfs, M. Munoz-Ibanez and A. G. Livingston, Nature Materials, 2016, 15, 760.

38. S. Lin and L. Yang, Tetrahedron Letters, 2005, 46, 4315-4319.

39. D. W. Tomlin, A. V. Fratini, M. Hunsaker and W. Wade Adams, Polymer, 2000, 41, 9003-9010.

40. S. Li, J. R. Fried, J. Colebrook and J. Burkhardt, Polymer, 2010, 51, 5640-5648.

41. M. Shan, X. Liu, X. Wang, I. Yarulina, B. Seoane, F. Kapteijn and J. Gascon, Science Advances, 2018, 4, eaau1698. 42. P. Totsatitpaisan, S. P. Nunes, K. Tashiro and S.

Chirachanchai, Solid State Ionics, 2009, 180, 738-745. 43. J. Yang and R. He, Polymers for Advanced Technologies,

2010, 21, 874-880.

44. S. Jenkins, K. I. Jacob, M. B. Polk, S. Kumar, T. D. Dang and F. E. Arnold, Macromolecules, 2000, 33, 8731-8738. 45. A. K. Sekizkardes, S. Altarawneh, Z. Kahveci, T. İslamoğlu

and H. M. El-Kaderi, Macromolecules, 2014, 47, 8328-8334.

46. A. K. Sekizkardes, J. T. Culp, T. Islamoglu, A. Marti, D. Hopkinson, C. Myers, H. M. El-Kaderi and H. B. Nulwala,

Chemical Communications, 2015, 51, 13393-13396.

47. S. C. Kumbharkar, P. B. Karadkar and U. K. Kharul, Journal

of Membrane Science, 2006, 286, 161-169.

48. M. G. Rabbani and H. M. El-Kaderi, Chemistry of

Materials, 2012, 24, 1511-1517.

49. A. Ebadi Amooghin, M. Omidkhah and A. Kargari, RSC

Advances, 2015, 5, 8552-8565.

50. T. C. Merkel, M. Zhou and R. W. Baker, Journal of

51. L. Giordano, J. Gubis, G. Bierman and F. Kapteijn, Journal

of Membrane Science, 2018, DOI: https://doi.org/10.1016/j.memsci.2018.12.063.